Next Article in Journal
Estimation of Actual Evapotranspiration and Its Components at Hourly and Daily Scales Using Dual Crop Coefficient Method for Water-Saving Irrigated Rice Paddy Field
Previous Article in Journal
The Effects of Different Crop Rotations on the Quality of Saline Soils in the Yinbei Plain
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 15
Issue 9
10.3390/agronomy15092132
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Use of a Digital Twin for Water Efficient Management in a Processing Tomato Commercial Farm

by
Sandra Millán
1,*,
Cristina Montesinos
1,
Jaume Casadesús
2,
Jose María Vadillo
1 and
Carlos Campillo
1
1
Centre for Scientific and Technological Research of Extremadura (CICYTEX), Agronomy of Horticultural Crops, Finca La Orden, Highway A-V, Km 372, Guadajira, 06187 Badajoz, Spain
2
Program of Efficient Use of Water in Agriculture, Institute of Agrifood Research and Technology (IRTA), Parc de Gardeny (PCiTAL), Fruitcentre, 25003 Lleida, Spain
*
Author to whom correspondence should be addressed.
Agronomy 2025, 15(9), 2132; https://doi.org/10.3390/agronomy15092132
Submission received: 8 August 2025 / Revised: 31 August 2025 / Accepted: 2 September 2025 / Published: 5 September 2025
(This article belongs to the Section Water Use and Irrigation)
Browse Figures
Versions Notes

Abstract

The increasing pressure on water resources caused by agricultural intensification, the rising food demand and climate change requires new irrigation strategies that improve the sustainability and efficiency of agricultural production. The objective of this study is to evaluate the performance of the digital twin (DT), Irri_DesK, in a 15-hectare commercial processing tomatoes plot in Extremadura (Spain) over two growing seasons (2023 and 2024). Three irrigation strategies were compared: conventional farmer management, management based on a remote-sensing platform (Smart4Crops) and automated scheduling using Irri_DesK DT-integrated soil moisture sensors, climate data and simulation models to adjust irrigation doses daily. Results showed that the DT-based strategy allowed for the application of regulated deficit irrigation strategies while maintaining productivity or fruit quality. In 2023, it achieved an economic water efficiency of 284.81 EUR/mm with a yield of 140 t/ha using 413 mm of water. In 2024, it maintained high production levels (126 t/ha) under more challenging conditions of spatial variability. These results support the potential of DTs for improving irrigation management in water-limited environments.

1. Introduction

Increased agricultural activity is putting rising pressure on available water resources, a situation exacerbated by the effects of climate change [1]. According to the United Nations World Water Development Report 2024 [2], agriculture accounts for approximately 70% of global freshwater withdrawals. In addition, there is a growing demand for food from an ever-increasing world population, which is projected to reach 9.8 billion by 2050 [3] and 11.2 billion by 2100 [4]. Such a scenario calls for the adoption of more efficient irrigation systems that optimize water use and help mitigate the effects of climate change.
Irrigation scheduling involves making decisions based on technical criteria regarding the timing and amount of water to be applied, taking into account the type of crop, its stage of development and environmental conditions. Traditionally, irrigation scheduling was based on empirical assessments of soil moisture by sensory perception, a qualitative method that relied heavily on the experience of the farmer [5]. Subsequently, the FAO-56 soil water balance (SWB) methodology was introduced [6] and has become the dominant approach for irrigation planning and management worldwide. The SWB is widely used to determine the irrigation requirements of a crop, balancing water inputs to the soil–plant system with expected yields. One of the key SWB components is the maximum crop evapotranspiration (ETc), which is the product of the reference evapotranspiration (ETo) and a crop coefficient (Kc) specific to each crop species and phenological stage. One of the methods used to estimate ETo is the Penman–Monteith method, which incorporates meteorological parameters such as temperature, relative humidity, solar radiation and wind speed, reflecting the demand imposed by meteorological conditions. For its part, the Kc is an adjustment factor that reflects the physiological characteristics of the crop and its variability throughout the development cycle [6]. However, the SWB method has certain limitations, as it does not consider possible deviations between estimated and actual water consumption due to specific site conditions, interannual variability in crop development [7], the plant material used, or the agronomic management practices applied [8]. These discrepancies can lead to systematic errors that tend to accumulate throughout the crop cycle, affecting the accuracy of irrigation scheduling.
In Spain, the region of Extremadura accounts for around 70% of the national processing tomato (Solanum lycopersicum L.) production [9]. In this region, this crop requires a lot of water, typically exceeding 600 mm per crop cycle [10]. Processing tomatoes are one of the crops most affected by irrigation restrictions due to climate change. As pointed out by Pérez et al. [11], the Extremadura Agricultural Law (Law 6/2015) establishes that, in water emergency situations, woody crops have priority for irrigation over annual crops, such as processing tomatoes. This reinforces the need to implement more efficient irrigation strategies.
The implementation of regulated deficit irrigation (RDI) strategies has been widely studied by the scientific community, demonstrating their effectiveness in saving water and improving quality in different crops [12,13,14,15,16,17]. RDI strategies consist of applying water deficits during the phenological stages least sensitive to water stress in order to limit vegetative growth with minimal impact on yield and fruit quality [18]. Water deficit usually causes a decrease in photosynthesis, plant growth and crop productivity, as well as beneficial effects on some fruit quality parameters, such as increased levels of antioxidant compounds and sugar accumulation [19]. In the case of processing tomatoes, the use of these RDI strategies helps to reduce transpiration. This can affect plant growth but, if applied at key moments, does not harm the crop [20]. In addition, the adoption of an RDI strategy in processing tomato increases the total soluble solids in the fruit, which is the main quality indicator used to determine the economic yield of the crop [21]. However, although there is a large body of scientific and technical knowledge on the best irrigation strategies for each crop type and growth stage, their application in the field remains a challenge. One of the main obstacles is the lack of adequate tools and the need for specific skills to accurately measure and monitor the water requirements of crops, thereby enabling the efficient distribution of water at the right time and place. To address this challenge, the development of digital twin (DT) technology can play a key role in the more efficient management of available water resources.
DTs are virtual models connected in real time to physical systems, enabling continuous and remote simulation, monitoring and optimization of agricultural processes [22]. DTs are increasingly being applied in agriculture as virtual replicas of real-world entities, supported by complex data analytics, predictive models and machine learning [23]. Ariesen-Verschuur et al. [24] conducted a comprehensive review of the implementation of DTs in greenhouse horticulture. Ahmed et al. [25] provided an overview of the application of artificial intelligence, deep learning and predictive models to optimize water use in agricultural areas affected by water scarcity, highlighting some cases of their use in greenhouses. Recent studies have demonstrated the application of DT technology in horticultural systems, highlighting its potential for real-time monitoring and decision-making in sensitive crops such as tomatoes [26], strawberries [27], rice [28] and corn [29]. Through various research projects, a DT called Irri_DesK has been successfully tested on different woody crops such as plum [30], apple [31], high-density olive groves [32] and vineyards [33]. Irri_DesK allows for automatic precision irrigation by calculating the water needs of the crops on a daily basis and adjusting this information with data from soil/plant sensors, meteorological sensors and remote sensing. Irri_DesK thus performs a daily closed loop to determine the irrigation dose to be applied to the crop, according to the strategy specific to the site, and automatically sends this irrigation dose to an irrigation programmer installed in the field. Irri_DesK also makes it possible to simulate individual SWB components, such as crop transpiration, soil evaporation and deep percolation, allowing farmers and technicians to see in real time the amount of water used or the water status of the crop. In addition, this DT has allowed the adoption of RDI strategies in plum [30] and high-density olive groves [32], improving water productivity compared to manual management by experts and reducing the time needed for irrigation monitoring and control by 80%. However, this DT has not been validated in open field horticultural crops.
In recent years, DT has played an increasingly important role in irrigated agriculture, particularly with regard to optimizing water use by promoting smart farming practices. Several studies have investigated using DT to integrate soil, climate and crop sensors to simulate irrigation strategies and optimize water use [22,34]. While studies on the use of DT in processing tomatoes are scarce, recent experiments have demonstrated the potential of combining artificial intelligence (AI), sensors and Internet of Things (IoT) platforms for programming irrigation. Martelli et al. [35] developed a decision support system (DSS) based on machine learning techniques for processing tomatoes under drip irrigation. This enabled the implementation of RDI strategies, resulting in significant water savings and improved fruit quality. Similarly, Galaverni et al. [36] developed an IoT platform for processing tomato crops that optimized irrigation, achieving water savings of up to 40%, while also paving the way for future advancements through the use of AI and DT.
Despite recent advances, studies on DTs in processing tomatoes are all very limited, and little evidence exists regarding their applicability in real commercial fields. The objective of this work is to verify the technical feasibility and evaluate the productive response in a commercial processing tomato plot located in Vegas Bajas del Guadiana, using the Irri_DesK tool to implement an automated drip irrigation system. In addition, an evaluation is made of how Irri_DesK achieves profitable production with a water consumption limit of less than 5500 m3/ha. Furthermore, using this DT for irrigation management will reduce the volume of water applied without compromising yield or fruit quality (ºBrix) while also increasing water use efficiency compared to conventional or remote sensing-based management.

2. Materials and Methods

2.1. Study Site

The study was carried out over two growing seasons (2023–2024) in a commercial plot located at the Aldea del Conde farm in the municipality of Talavera la Real (Badajoz) (latitude 38°84′65.64′ N, longitude 6°72′59.18′ W, datum WGS84) (Figure 1a). The plot covers an area of 15 ha of processing tomatoes (Lycopersicum esculentum Mill). In the 2023 season, the UG16112 variety was transplanted from 6 to 10 April, and in 2024 the H1015 variety from 15 to 19 April, following a west–east direction on the plot. The planting density is 29,500 plants per hectare, planted in double rows. The plants are staggered, with a distance of 46 cm between successive plants. The soil is classified texturally as sandy loam, with an average clay, silt and sand content of 12.89%, 17.02% and 70.08%, respectively. It has a near-neutral pH level of 6.5–7.9 and an average organic matter content of 0.7–1.5%. The effective rooting depth is around 1 m, and the terrain is flat with a slope of 0%. In terms of management history, crop rotation is commonly practiced on the plot, alternating processing tomatoes with other extensive crops. In the year prior to the trial (2022), the plot was left fallow, and, exceptionally, processing tomatoes were grown in two consecutive seasons (2023 and 2024) for the purpose of conducting the research. The climate of the area is Mediterranean with a slight Atlantic influence. The dry season is from June to September (summer) and the wet season is from October to May (winter), during which about 80% of the annual rainfall is concentrated. For the period 2010–2024, the average ETo and precipitation values were 1303.56 mm and 473.65 mm, respectively. During the same period, the average maximum and minimum air temperatures were 23.89 °C and 9.80 °C, respectively. The warmest months are July and August, with maximum temperatures above 40 °C almost every year, although they rarely exceed 45 °C. The coldest months are December and January, with annual minimum temperatures below 0 °C and extreme values rarely below −5 °C. The plot was divided into three sectors according to the irrigation management applied (A, B and C) with areas of 5.5 ha, 5.12 ha and 4.34 ha, respectively (Figure 1b). Fertilization and pest and disease control of the processing tomatoes were carried out by technicians from the Aldea del Conde farm. The tomatoes were irrigated daily using a drip irrigation system with 1.05 L/h emitters spaced 0.30 m apart.

2.2. Characterization of the Spatial Variability of the Plot, Selection of Control and Monitoring Points

Prior to transplanting, the Dualem-1S sensor (Dualem, Inc., Milton, ON, Canada) was used to characterize the spatial variability of the plot based on the apparent electrical conductivity (ECa) of the soil (Figure 1c). The Dualem-1S sensor was equipped with a global positioning system and operated at a frequency of 9 kHz. This device has one transmitter (Tx) and two receivers (Rx) spaced 1 m apart, allowing the ECa to be measured at depths of 0–0.50 m and 0–1.50 m. For data acquisition, the sensor was mounted inside a 3 m long polyvinyl chloride structure that was transported by a moving vehicle at an average speed of 9 km h−1. Due to the height of the structure, ECa measurements were recorded in the 0–0.40 m and 0–1.40 m layers, assuming a zero value for air ECa.
In this study, measurements corresponding to the surface layer (0–0.40 m) were analyzed as the greatest root activity of tomatoes is found in the first 0.30 m of the soil profile. Data were collected on 5 April 2023, following parallel transects approximately 4 m apart, with a recording frequency of one measurement per second.
A total of 2482 ECa measurements were obtained, which were interpolated using ordinary kriging with the geostatistical software QGIS 3.34.11 (QGIS Development Team, 2024; https://www.qgis.org).
The ECa was related to soil texture and water content to identify management zones within the plot. The map was made in April when the soil water content (SWC) was mainly derived from recorded rainfall. Three different zones were distinguished on this ECa map (Figure 1d): (I) green, lighter soil zone, with lower ECa values (5–10 mS/m), a higher sand content and a lower water holding capacity; (II) yellow, with intermediate ECa values (11–19 mS/m), an intermediate soil texture and water holding capacity; and (III) white, zone with higher ECa values (20–34 mS/m), with a more clayey soil texture and a higher water holding capacity. These differences in ECa values are due to a complex interaction of biological, agronomic, edaphic, anthropogenic, topographic and climatic factors [37]. In 2023, the DT of Irri_DesK was implemented in Zone B, while in 2024 it was used to manage irrigation in Zone A. In both years, seven control points (CPs) were established, and nine additional monitoring points (MPs) were added to the CPs (Figure 1d,e). The CPs and MPs were selected using the Auravant web platform (https://www.auravant.com).

2.3. Analysis of Soil Properties and Determination of the Water Content in the Soil at Field Capacity, the Permanent Wilting Point and the Water Available to the Plant

Table 1 presents the soil properties in the study area. A soil sample was collected at each CP and MP on 27 March 2023, at a depth of 0.30 m. All samples were transported to the laboratory in plastic bags, air-dried, crushed and sieved to 2 mm. Soil characterization included the determination of texture, pH and organic matter (OM) content. Texture was analyzed using the hydrometer method [38], pH was measured in a 1:2.5 (soil–water) suspension using the potentiometric method [39], and OM content was determined VIA dichromate oxidation [40].
In addition, an unaltered soil sample was taken at the CPs to determine the SWC at field capacity (FC) and at the permanent wilting point (PWP), the SWC available to the plant (WAP), and the apparent bulk density. The HYPROP method (UMS, Munich, Germany) was used, following the methodology described by Schindler et al. [41], in combination with the WP4C dew point potentiometer (Decagon Devices, Pullman, WA, USA). The sample was collected on 21 November 2022 using a 250 cm3 diameter stainless steel ring at a depth of 0.2 m. It was brought to capillary saturation in the laboratory and then analyzed using HYPROP according to the manufacturer’s instructions [42]. HYPROP is a device based on the simplified evaporation method [43], which allows direct measurement of soil water potential by inserting small tensiometers into the sample while the moisture content gradually evaporates. The instrument operates in a suction range of 0–85 kPa. A disturbed soil sample was also taken in the same points at a depth of 0.2 m for analysis with a WP4C dew point hygrometer (Table 2), which measures water potential in the range −0.1 MPa to −300 MPa. For analysis with the WP4C, samples were sieved and moistened to varying degrees [44]. The combination of data obtained with HYPROP and WP4C allowed the determination of soil water content values at FC, PWP and WAP using the HYPROP-FIT software (version 4.2.2, METER Group) adapted to different models (Table 2)

2.4. Automatic Irrigation System and Definition of Irrigation Seasonal Plan

The automatic irrigation system implemented in this study consisted of (a) sensors installed in the field (Figure 2) and (b) a DT called Irri_DesK, used for irrigation scheduling.

2.4.1. Local Sensors Installed in the Field

At each control point, the SWC was monitored using three Teros 10 capacitance probes (Decagon Devices, Inc., Pullman, WA, USA), using the factory calibration supplied by the manufacturer, which is recommended for mineral soils. No additional field calibration was performed. To characterize the wet bulb in the drip influence area, three sensors were installed at a depth of 0.20 m and 0.10 m from the drip in a horizontal direction towards the bed (Figure 2). In addition, an MTKD water meter (LabFerrer S.L., Cervera, Lleida, Spain) was installed at the different CPs and MPs to measure the amount of water applied and to record the irrigation events. An Apogee SI411 thermal sensor (Decagon Devices, Inc., Pullman, WA, USA) was also installed at each CP. All sensors were wired to a ZL6 data logger (METER Group, Pullman, WA, USA) and data were stored at 30 min intervals on the ZENTRA cloud platform (https://zentracloud.com, accessed on 7 August 2025). Irri_DesK downloaded the data stored on the ZENTRA Cloud web platform daily at 3:00 am.

2.4.2. Digital Twin (Irri_DesK)

Irri_DesK (www.irridesk01.com, accessed on 7 August 2025) is a web-based decision support tool that integrates crop specific information, sensor data, meteorological inputs and simulation models [33] to estimate daily irrigation requirements. In this study, the platform was configured to determine daily doses for each management zone [8]. While the system is capable of direct communication with irrigation controllers, in this case recommendations were manually implemented by a technician based on daily messages sent via WhatsApp. The technician then manually entered the recommended dose into the irrigation controller.
Prior to the irrigation campaign, three main configuration steps were carried out:
  • Description of the plot: This section provides detailed information on the characteristics of the crop and the characteristics of the soil. The selected crop is processing tomato, with a maximum root depth of 0.4 m and a maximum vigor of 0.8 on a scale of 0 to 1, expressed in terms of soil cover or intercepted radiation (Figure 3a). This crop was established without vegetation cover (Figure 3a). The soil on the plot is texturally classified as sandy loam according to the USDA classification, with a slope of 0% and a depth of 1 m. In 2023, the SWC was 0.254 m3 m−3 at FC and 0.055 m3 m−3 at PWP (average of the values obtained at CP-3, CP-4 and CP-5). In 2024, the SWC was 0.245 m3 m−3 at FC and 0.057 m3 m−3 at PWP, calculated as the average of the values obtained at CP-1 and CP-2. A drip irrigation system was used with a drip spacing of 0.30 m, a nominal flow rate of 1.05 L/h per drip and a wet zone diameter of 0.45 m. The distance between crop beds was 1.5 m.
  • Sensor configuration: In this section, Irri_DesK must be provided with the list of sensors to be integrated into the automatic irrigation system. In this study, the information from the sensors previously installed in the field was entered (Figure 2). The meteorological data were obtained through the application program interface (API) of the Agroclimatic Information System for Irrigation (SiAR by its initials in Spanish), which provides real-time information. The data correspond to a weather station (Campbell Scientific, Logan, UT, USA) located 6.5 km from the study plot, managed by the Centre for Scientific and Technological Research of Extremadura, part of the Regional Government of Extremadura. Effective precipitation (Pe) was estimated using the proposed method of the FAO, which defines this parameter as the monthly precipitation exceeded in 80% of the analyzed years [45].
  • Seasonal schedule configuration: A key feature of Irri_DesK is irrigation scheduling based on a seasonal plan, which facilitates the application of more advanced strategies such as RDI, a solution that is particularly useful for farmers with water restrictions. In this seasonal plan, the amount of accumulated water in millimeters that will be used by the crop during the irrigation season is estimated. In the case of the processing tomato analyzed, the seasonal plan was provided by the farmer with a value of 550 mm. It is also necessary to set an upper limit (maximum cumulative irrigation allowed in an irrigation season) and a lower limit (minimum reasonable cumulative irrigation in an irrigation season). The water thresholds used as upper and lower limits were set at 650 mm and 400 mm, respectively (Figure 3b). When measured irrigation approaches the upper or lower limit of the established range, the irrigation doses are calculated so that cumulative irrigation remains within the allowed values. In addition, water imbalances can be established at specific times during the irrigation campaign, expressed as the ratio between the irrigation that should be applied and that estimated by the SWB model. In the seasonal plan, Irri_DesK was instructed to apply a water imbalance during the crop’s ripening phase [20]. Since the objective was to implement an RDI strategy, water status levels were defined based on a multiplier coefficient applied to irrigation requirements or estimated deviations from the SWB. So, during the ripening period, a reduction factor of 0.5 was set to stress the crop at this phenological stage (Figure 3c). The seasonal irrigation plan was drawn up based on simulations carried out using the SWB model. This involved taking detailed information into account on crop and soil characteristics, the irrigation system used, the plot’s ten-year meteorological history, the range of irrigation applied in previous campaigns and the intended deviation curve of the SWB.
Irri_DesK is available in a free, non-commercial version for research purposes and in a commercial version distributed by an external company. The research version (v.2023) of the Irri_DesK DT was used for irrigation control (Figure 4).

2.5. Irrigation Scheduling in the Different Management Zones

The plot was subdivided into three zones with different drip irrigation management criteria, as follows:
In 2023 (Figure 1d):
  • Zone A: Irrigation management was carried out by technicians from a precision agriculture company (E_AP), using information provided by the Smart4Crops platform. This platform used multispectral satellite imagery (Sentinel-2) to estimate crop vigor (NDVI). A technician adjusted the timing and volumes of irrigation weekly based on vegetation indices, without integrating soil moisture data or in-field sensors.
  • Zone B: Irrigation management was carried out following the information provided by Irri_DesK. The system applied full irrigation during sensitive phenological stages (transplant, vegetative growth and fruit development) and introduced a 50% reduction (RDI) during ripening. In Irri_DesK, irrigation doses were determined daily using a soil water balance model adjusted with in-field soil moisture sensors and meteorological inputs (ETo from the SiAR station).
  • Zone C: Irrigation was managed conventionally by the farmer. In this zone, irrigation was scheduled to cover crop water needs throughout the crop cycle.
In 2024 (Figure 1e), zone B was managed by the precision agriculture company and the Irri_DesK DT was implemented in zone A due to its high spatial variability. This selection allowed the performance of the Irri_DesK tool to be evaluated in a more unfavorable environment.

2.6. Agronomic Measurements

2.6.1. Normalized Difference Vegetation Index (NDVI)

At each CP and MP, the normalized difference vegetation index (NDVI) was analyzed using the Auravant web platform (https://www.auravant.com). This tool uses satellite images from the Sentinel-2 sensor (Copernicus program, European Space Agency) with a spatial resolution of 10 m and an update frequency of 3 to 5 days under favorable atmospheric conditions. Only images with a maximum cloud cover of 20% were considered to ensure that the CP and MP areas were free from cloud contamination. For each analysis date, the NDVI values corresponding to each CP and MP were extracted and the average of these values was then calculated. The NDVI is a widely used indicator for assessing vegetation development and vigor, as it allows differences in canopy cover within a field to be detected [46]. The NDVI was calculated from the near-infrared (B8) and red (B4) spectral bands using the formula:
NDVI = (B8 − B4)/(B8 + B4)

2.6.2. Yield and Quality

Commercial production, total production and quality of industrial tomatoes were determined at both the CPs and MPs. Prior to harvest, five additional points were selected in each zone, taking into account the spatial variability present in order to obtain an adequate representation of crop performance. In total, 31 sampling points were harvested across all plots and years. Harvesting took place on 25 July 2023 and 29 July 2024, coinciding with the time when more than 85% of the fruits were red. At each time point, 24 plants of 9 m2 were harvested. The fruits were divided into two categories: commercial production, corresponding to red tomatoes of market quality, and total production, calculated as the sum of all fruits harvested (red, green and discarded).
To determine quality, 30 red fruits were randomly selected at each sampling point and the total soluble solids content (°Brix) was measured using a digital refractometer (Mettler Toledo, model RE4OD, Columbus, OH, USA).

3. Results and Discussion

3.1. Exploratory Analysis of the ECa Data

Table 3 shows the descriptive statistics of the ECa in the study plot at a depth of 0–0.04 m. According to Goovaerts [47], although the assumption of normality is not strictly necessary for the application of kriging, it is a desirable characteristic. This is because if the random variable follows a normal distribution, kriging provides optimal estimates in terms of minimum variance. The ECa values obtained in the surface soil layer reflect moderate variability in soil properties within the study area. The similarity between the mean and the median, as well as the low skewness value, suggests that the data distribution is approximately symmetric, with no significant presence of extreme values. However, the wide ECa range (2 to 47 mS/m) reveals significant differences in soil conditions, which could be related to variations in soil texture, moisture content or the accumulation of soluble salts. The slight platykurtosis observed (kurtosis < 3) implies a lower presence of extreme values compared to a normal distribution, which may indicate a more homogeneous distribution of ECa in the analyzed profile. These results support the use of ECa as an effective tool for characterizing soil heterogeneity, which is relevant for the design of site-specific agronomic management strategies or precision irrigation systems.
In the second phase of the geostatistical study, the spatial distribution of ECa was quantified using a variogram. The model that best fitted the experimental variogram was a stable model (Figure 5) with the following parameters: range = 355 m, sill = 0.005 and nugget = 0.001. As a result, the ratio between the nugget effect and the sill was 20%, indicating a strong spatial dependence, as this ratio was close to 25% which is the commonly accepted threshold for classifying the dependence as strong. According to Cambardella et al. [48], spatial dependence can be classified as follows: strong (nugget/threshold ratio ≤ 25%), moderate (nugget/threshold ratio between 25% and 75%) and weak (nugget/threshold ratio > 75%). In this case, the low proportion of the seed effect suggests that most of the spatial variability of ECa can be explained by correlated spatial structures, which is favorable for more accurate interpolations using kriging.
Validation of the kriging model was carried out by evaluating the performance metrics obtained from the prediction errors (Table 4). The mean error (ME), root mean square error (RMSE) and mean deviation (MD) were calculated for the depth interval of 0–0.40 m. The ME, RMSE and MD showed values close to zero, indicating minimal bias and satisfactory agreement between observed and predicted ECa values. These results confirm that the model provides accurate and reliable interpolation for the study area, supporting its application in similar soil variability assessments.

3.2. Applied Water in the Different Management Zones

Figure 6 shows the amount of water applied per irrigation, the total cumulative rainfall and the effective cumulative rainfall in the three management areas (A, B and C) during the 2023 and 2024 seasons. According to Duncan’s test, there were significant differences in the volume of irrigation applied between treatments in 2023 (p < 0.05). The conventional Farmer treatment had the highest application rate, followed by Smart4Crops. In contrast, the Irri_DesK DT applied significantly less water, achieving a water saving of 26% compared to the Farmer treatment and 13% compared to the Smart4Crops platform. These water savings can be partially attributed to the ability of the DT to respond rapidly to changes in soil and environmental conditions. Casadesús et al. [8] employed a water balance model along with dynamic adjustments to tailor the amount of water applied to the specific needs of each plot. In their study, soil moisture sensors integrated into the system provided the precision required for this adjustment. Tancredi et al. [49] applied a DT with machine learning algorithms to an olive crop grown in a greenhouse, achieving precise control of drip irrigation and high decision accuracy.
In 2024, the Irri_DesK DT was implemented in Zone A, which presented greater soil heterogeneity. The system adjusted irrigation volumes slightly upward compared to 2023, probably in response to a higher water demand associated with local soil and climatic conditions. However, the applied water remained within the range defined in the seasonal irrigation plan (400–650 mm), suggesting that the DT adapted irrigation to prevailing field conditions without significant over- or under-irrigation. In zone C, managed by the farmer, there were no significant differences in the volume of water applied with the Irri_DesK DT. This coincidence was not accidental, as the farmer decided to replicate the irrigation strategy suggested by Irri_DesK that year. This highlights the value of the DT, not only as an automation tool, but also as a source of transferable knowledge to improve traditional practices. The treatment managed by the precision agriculture company applied significantly less water than the other treatments.
Rainfall varied across the two years, with higher total and effective precipitation in 2023 than in 2024. This variability affected the relative contribution of irrigation versus rainfall to total crop water supply, emphasizing the importance of considering effective precipitation in the water balance to avoid overestimation.
Figure 7 shows the evolution of the amount of water applied during the growing cycle of industrial tomatoes in the area managed by the DT (Irri_DesK), compared to the pre-established seasonal irrigation plan.
In 2023 (Figure 7a), during the transplanting phase (Phase I), Irri_DesK applied a significantly lower amount of water than defined in the seasonal plan. At the time of transplanting, the soil profile is fully saturated due to recent rainfall or pre-sowing irrigation, allowing irrigation to be reduced without compromising development. This strategy avoids over-irrigation at a time when the crop’s water requirements are low [20] and can be adapted to the available soil moisture. This real-time adjustment capability reflects one of the key advantages of Irri_DesK over static systems, namely integration of moisture sensors, water balance and climate data. During the rapid vegetative growth phase (Phase II), which is characterized by accelerated vegetative growth and a high water demand, Irri_DesK closely followed the seasonal schedule, avoiding any water deficit during this period so as not to cause stress to the plant that could lead to flower abortion. In the fruit development phase (Phase III), which corresponds to fruit enlargement, precise watering is also required. Here Irri_DesK maintained its behavior according to the seasonal plan, ensuring sufficient water availability to maximize fruit size without compromising quality. In Phases II and III, Irri_DesK closely followed the seasonal plan without implementing RDI, as these phases are particularly sensitive to water stress. As previous studies on processing tomatoes have shown [19,20], water shortages in these phases can cause flower abortion, fruit size reduction and yield reduction. The behavior observed in Irri_DesK shows that its algorithm recognizes these phases as critical and maintains full irrigation in accordance with agronomic recommendations. Finally, in the ripening phase (Phase IV), when the fruit begins to ripen (transition from green to red), the system applied RDI strategies, reducing the irrigation volume with respect to the seasonal plan. Mild stress during this phase favors the concentration of soluble solids (ºBrix), improves fruit quality and does not affect commercial production [20,21].
In 2024, Irri_DesK was applied in an area with greater spatial heterogeneity. Figure 7b shows a similar trend to 2023, with less water applied in Phase I due to the initial soil moisture. In Phases II and III, the seasonal schedule was again strictly followed. However, a slight increase in the slope of the accumulation curve is observed compared to 2023, indicating that the system responded to higher water demands, possibly associated with lower effective precipitation and higher heat stress. In Phase IV, RDI strategies were again implemented, but in a more restrained manner than in the previous year. This suggests that the system adapted to the soil and climate conditions of the 2024 season, seeking to maintain water efficiency without compromising quality or yield. This ability to apply RDI in a precise and automated manner demonstrates the potential of DTs as a tool for optimizing water use more efficiently. Similarly, Millán et al. [32] demonstrated that the DT was able to implement an appropriate RDI strategy for a hedge olive grove located in the areas of the plot with the most unfavorable conditions, characterized by high vegetative vigor and low ECa.

3.3. Soil Water Content

Figure 8 shows the evolution of soil water content (θw) during the 2023 season for the different irrigation zones (A, B and C). In zone A, managed using the Smart4Crops platform, control points CP-1 and CP-2 showed differences in their FC and WAP thresholds, attributable to variability in soil physical properties such as texture and bulk density [50,51]. During Phase I, all sensors remained within the optimal range (FC–WAP), reflecting adequate initial moisture. However, slight differences were observed between sensors at CP-1, such as S1, which recorded lower values (0.164 m3/m3) than S2 and S3, possibly due to its location relative to the emitter. These differences in SWC could be associated with local thermal variations due to water evaporation from the soil surface or changes in the thermal conductivity of the soil, which is directly influenced by its moisture level [52,53]. In Phase II, the behavior was more variable. At CP-1, S1 and S2 showed values above FC, while S3 fell towards WAP. In CP-2, all sensors temporarily dropped below WAP in early May, indicating episodes of moderate stress, although S2 subsequently recovered to adequate levels. This variability highlights the need to use multiple sensors at the same depth, located in strategic positions with high root activity, in order to improve the spatial representativeness of measurements within the wet bulb and increase the reliability of the recorded data [31,54]. In Phase II, behavior was more variable. At CP-1, S1 and S2 showed values above FC, while S3 fell towards WAP. In CP-2, all sensors temporarily fell below WAP in early May, indicating episodes of moderate stress, although S2 subsequently recovered to adequate levels. This variability highlights the need to use multiple sensors at the same depth, located in strategic positions with high root activity, in order to improve the spatial representativeness of measurements within the wet bulb and increase the reliability of the recorded data [31,54]. Furthermore, in localized irrigation systems, this variability is more relevant, as wet bulbs form around the emitters, while the rest of the soil may be only slightly affected or not affected at all by irrigation. In the case of horticultural crops, this variability may be greater, given that moisture sensors must be located within the wet bulb, an area where root density is usually higher [8]. Kang et al. [55] indicated that root size can affect sensor readings. During Phase III, marked oscillations were observed between irrigations. At CP-1, S1 remained above FC, while S2 and S3 gradually fell towards the FC–WAP range. At CP-2, S1 recorded values below the WAP threshold, reaching a minimum of 0.156 m3/m3. In Phase IV, at CP-1, S1 showed a progressive decrease, reaching values close to WAP at the end of the cycle, while S2 and S3 remained within the optimal range. At CP-2, S2 remained slightly above FC throughout the phase, with a maximum value of 0.273 m3/m3. In zone B, managed by the DT Irri_DesK, more stable soil moisture behavior was observed. In Phase I, the sensors remained mostly between FC and WAP, except for sensor S3 at CP-3, which fell slightly below WAP (minimum of 0.183 m3/m3). In Phase II, moisture remained within optimal ranges, although at CP-5, S1 and S2 temporarily exceeded FC (0.314 and 0.260 m3/m3), probably due to their proximity to the emitter. In Phase III, Irri_DesK prevented deficits, although S3 at CP-3 fell below WAP towards the end of the stage (minimum of 0.124 m3/m3). In Phase IV, an RDI strategy was applied, resulting in a progressive reduction in SWC. The sensors recorded a moderate decrease to values close to WAP, without reaching severe stress conditions, except for sensor S2 located at CP-4 (which reached a minimum value of 0.126 m3/m3). This behavior was due to the agronomic strategy applied, aimed at improving fruit quality (increase in ºBrix) without compromising production [20].
Finally, in zone C (conventional management), during Phase I, all sensors remained between FC and WAP, except for some peaks above FC in sensors S1 at CP-6 and CP-7. In Phase II, more pronounced oscillations were observed. For example, S1 at CP-6 fell below WAP (minimum of 0.154 m3/m3), while other sensors showed an oversupply. During Phase III, S1 at CP-6 remained below WAP for most of the period (minimum of 0.109 m3/m3), suggesting an underestimation of irrigation. In Phase IV, most sensors indicated excess irrigation (values above FC), except again for S1 at CP-6, which remained below WAP throughout the phase (minimum of 0.101 m3/m3 at the end of the irrigation campaign).

3.4. Agronomic Response and Water Use Efficiency

Figure 9 shows the temporal evolution of the NDVI in the three management areas during the 2023 (Figure 9a) and 2024 (Figure 9b) irrigation seasons. The NDVI is a widely used indicator of vegetative vigor, crop cover and, indirectly, crop physiological status. It is commonly used to monitor crop growth and analyze crop response to water conditions and agronomic management practices [46]. For the processing tomato crop, an optimal physiological state without water limitations is associated with NDVI values in the range of 0.70 to 0.85, especially during the phases of maximum vegetative development [56]. However, greater canopy development does not necessarily imply higher productivity. Pettorelli et al. [57], it was found that temporary suspensions of irrigation, even with partial recovery of the leaf area index (LAI), reduced commercial yield by up to 33%, demonstrating that the amount of vegetative biomass does not always translate into higher fruit production.
In 2023 (Figure 9a), NDVI trends varied across management zones. Zone B, managed using the Irri_DesK DT, reached higher NDVI values and exhibited a more consistent increase over time, suggesting more uniform crop development. Zone A, managed using the Smart4Crops platform, showed intermediate NDVI levels, while zone C, under conventional management, had lower NDVI values and greater temporal variability. These patterns may reflect differences in irrigation timing and volume relative to crop water needs, particularly during sensitive growth phases. In Phase I, NDVI values were low (below 0.30) across all areas due to the initial poor crop coverage. However, the area managed using the Smart4Crops platform showed higher values than the others, as it was the first to be transplanted. During Phase II, the NDVI increased exponentially to values close to 0.85–0.9, reflecting rapid crop development. During this period, the area managed by Irri_DesK recorded the highest NDVI values. This phase is also characterized by very active root growth, meaning that by the end of the phase, the roots have usually reached their maximum depth. In most cropland soils in the Guadiana Vegas (Extremadura, Spain), the greatest root volume is found within the first 0.30 m of soil depth [20]. Next, in Phase III, the NDVI values stabilized within the range of 0.70–0.85, which corresponds to the maximum level of crop vigor. A water deficit at this stage can have a negative impact on fruit set and ripening [20,21]. In phase IV, a slight decrease in the NDVI was observed, which was associated with a progressive reduction in crop vigor. During this phase, the area managed by Irri_DesK exhibited lower NDVI values, likely due to the application of RDI during this period [19,20,21].
In 2024 (Figure 9b), the DT was implemented in zone A, which had greater spatial heterogeneity. NDVI values in this zone were higher than those recorded in 2023, indicating a possible improvement in crop vigor. Zone C, again managed by the farmer, showed a similar NDVI evolution to 2023, but with a slight improvement. Notably, the farmer reported having adjusted irrigation practices based on observations from the previous season, potentially influencing these results. Zone B, now managed by the precision farming company, showed a more stable trend than in the previous season, although without reaching the NDVI peaks observed in the zone with Irri_DesK. These results suggest that irrigation strategies incorporating sensor data may support more consistent vegetative development, especially in zones with high spatial variability. As in 2023, the NDVI values in 2024 were initially low during Phase I (establishment), rising sharply to maximum values during Phase II (vegetative growth). They remained high and stable during Phase III (fruit development) and then declined during Phase IV (ripening) due to natural canopy senescence. Similarly, crop development in the Irri_DesK managed area is currently lower than in other areas. This could be due to the implementation of RDI strategies [19,20,21].
Figure 10 shows the total production, commercial production and total soluble solids content (ºBrix) for the different irrigation management strategies across the study zones.
In both 2023 and 2024, the Irri_DesK DT (zones B in 2023 and A in 2024) consistently resulted in a favorable balance between yield and fruit quality, demonstrating high efficiency in water use. The zones managed by the Irri_DesK DT achieved total production of 140 t/ha (2023) and 126 t/ha (2024), with ºBrix values of 5.42 and 5.82, respectively, using optimized water volumes (Figure 6). This agronomic response suggests an efficient water management adapted to the phenological stages of the crop, applying RDI strategies in less sensitive stages, such as ripening, without affecting yield [20,21]. These results highlight the capacity of sensor-based dynamic scheduling to adapt irrigation to soil water availability and crop phenology [32].
In contrast, conventional management (zone C in both years) reached the highest total production (148.59 t/ha in 2023 and 150 t/ha in 2024), but with varying fruit quality: ºBrix improved from 5.15 in 2023 to 5.68 in 2024, likely due to the adoption of partial DT strategies by the farmer.
The Smart4Crops platform (zone A in 2023 and zone B in 2024) showed contrasting results. In 2023, it produced the lowest yield (106.95 t/ha) but the highest ºBrix (5.53), possibly due to under-irrigation in heterogeneous soils since the system did not have on-site sensors [8]. However, when applied in zone B in 2024, Smart4Crops achieved a better result with higher yield (135 t/ha) and ºBrix (5.93).
The strategy applied with Irri_DesK included an RDI period during the ripening phase. In previous studies on processing tomatoes, this management has been associated with a greater accumulation of soluble solids [19,21] without negative effects on harvest uniformity. In our case, the Brix values remained within adequate ranges and without yield reductions, suggesting that ripening and harvesting uniformity were not compromised. However, this study did not evaluate in detail the soluble solids profile or harvest synchrony, aspects that should be addressed in future research to confirm and expand on these results.
Table 5 shows the income (EUR/t and EUR /ha), the amount of water applied per hectare and the economic water efficiency (EWE). In both campaigns, the area managed by the farmer achieved the highest gross income per hectare, with 23626 EUR/ha in 2023 and 25613 EUR/ha in 2024. However, this came at the cost of higher water application (128.80 mm/ha in 2023 and 115.67 mm/ha in 2024), reflecting lower water efficiency. In contrast, Irri_DesK, with lower allocations (80.66 mm/ha in 2023 and 94.36 mm/ha in 2024), achieved revenues of 22974 EUR and 21772 EUR/ha, respectively, while reducing water use by more than 30%. This result shows that through strategies such as RDI it is possible to maintain profitability without applying the maximum amount of water, especially when fruit quality is improved by increasing °Brix [19,20,21]. One of the most relevant indicators in water scarcity scenarios is the EWE (EUR of income per mm of water applied). This parameter allows assessment of the profitability per unit of the applied water resource, which is crucial for agronomic sustainability in areas affected by climate change. The highest values were achieved in 2023 with the use of Irri_DesK (284.81 EUR/mm) and in 2024 by Smart4Crops (279.01 EUR/mm 2024). This shows that advanced irrigation management technologies not only optimize the resource but also increase the economic return per unit of water used, which is essential for coping with campaigns under restrictive conditions [25,32]. In the current context of climate change, with agricultural seasons characterized by more frequent and intense droughts, the use of irrigation strategies such as those implemented by Irri_DesK is particularly relevant. As reported in Pérez et al. [11], the Agricultural Law of Extremadura (Law 6/2015) establishes that in the event of a water emergency, woody crops have priority for irrigation over annual crops such as processing tomatoes. However, even with this priority, available supplies may be limited. Under these conditions, tools such as Irri_DesK allow the cultivation of crops to continue, albeit in a smaller area, because the system optimizes resource use and increases profitability per unit of water. For example, with a maximum allocation of 5000 m3/ha, it would only be possible to maintain areas such as those managed with the efficient treatments (Irri_DesK or Smart4Crops). Conversely, conventional strategies would require a reduction in area of up to 25% to stay within the limits set. This reality reinforces the need for smart systems that maximize profitability per cubic meter of water applied, which is key in an environment with lower water availability [58].
In 2023, Irri_DesK applied lower irrigation volumes than the farmer, resulting in a slightly lower yield (140 vs. 148 t/ha), but with higher EWE. Thus, the system did not maximize absolute yield, but it optimized the ratio between water applied and yield obtained, achieving the highest profitability per unit of water. In 2024, under more restrictive conditions with lower rainfall and greater soil heterogeneity, Irri_DesK maintained competitive yield levels (126 t/ha), which demonstrates its robustness in more challenging environments. This behavior reflects previous studies reporting that precision irrigation systems tend to maximize efficiency under water-limited conditions [32,33], while absolute yields may be higher when irrigation is managed with greater water inputs [14,15,16,17].

3.5. Limitations of the Irrigation Management Methods

Although the use of the Irri_DesK DT improved water use efficiency and maintained adequate crop quality, its implementation requires an initial investment in sensors, licenses and data loggers, as well as digital connectivity and technical training, which may limit adoption on farms with fewer resources. In particular, the system depends on a stable internet connection for real-time data transmission, which can be a challenge in rural areas with limited coverage [24,25]. In addition, the initial costs of equipment can be significant and the need for specialized training to configure, interpret and maintain the platform may hinder its adoption by farmers with less experience in digital tools [22,23]. Furthermore, its performance depends on the availability of reliable climatic and soil data. Another important challenge is the spatial variability of soil moisture sensors, particularly under localized irrigation systems. Sensors installed in the same location may show similar patterns; however, sensor-to-sensor variability can arise due to differences in root distribution [55], wet bulb formation and soil heterogeneity [59]. These factors highlight the need for careful interpretation of sensor data to ensure accurate irrigation decisions.
The Smart4Crops platform, based on satellite imagery, offers a broad spatial overview of crops. However, the lack of real-time soil moisture data limits its adaptability in fields with highly heterogeneous soil types or under unfavorable weather conditions (prolonged cloud cover).
Conventional management, based on farmer experience, can be effective under certain conditions, such as in homogeneous soils, with stable climate patterns, or when managing familiar crops on small plots. However, it is more prone to overwatering or underwatering due to the lack of monitoring systems that provide information to aid decision-making.

4. Conclusions

The implementation of the Irri_DesK DT in a commercial processing tomato field confirmed its feasibility as a decision-support tool for irrigation under variable soil and climatic conditions. By integrating real-time data from soil sensors, weather sources and crop models, it enabled dynamic irrigation scheduling, including the application of RDI strategies during less sensitive growth stages.
Compared to conventional practices and remote sensing-based systems, the DT achieved greater water use efficiency, using less water while maintaining or improving agronomic outcomes. In 2023, it reached an optimal balance between yield (140 t/ha), quality (5.42 °Brix) and irrigation volume (413 mm), resulting in the highest economic return per millimeter of water applied (284.81 EUR/mm). In 2024, it maintained high productivity despite more challenging climatic conditions and soil variability, demonstrating robustness and adaptability. Additionally, water use remained within the planned seasonal range (400–650 mm), indicating effective adjustment to field conditions without significant over- or under-irrigation.
The use of the DT also facilitated knowledge transfer to manual irrigation practices, as farmer-managed plots that followed its recommendations showed improvements in fruit quality and irrigation efficiency.
Future research should focus on integrating ECa measurements into Irri_DesK as a criterion for delineating differential management zones, which would allow the irrigation algorithm to be adapted according to soil texture and water retention capacity. Incorporating generative artificial intelligence could further enhance the system’s capabilities by simulating and optimizing irrigation strategies tailored to both crop status and projected climate scenarios. Additionally, blockchain technology could offer new opportunities for traceability and access to sustainability incentives.

Author Contributions

Conceptualization, S.M. and C.C.; methodology, S.M. and J.M.V.; software, J.C.; validation, S.M. and C.M.; formal analysis, S.M. and C.M.; investigation, S.M. and J.M.V.; resources, C.C.; data curation, S.M. and C.M.; writing—original draft preparation, S.M.; writing—review and editing, C.M., C.C., J.C. and J.M.V.; visualization, S.M.; supervision, S.M.; project administration, S.M.; funding acquisition, C.C. All authors have read and agreed to the published version of the manuscript.

Funding

This work was developed in the ET4DROUGHT Project (PID2021-127345OR-C33) funded by the Spanish Ministry of Science and Innovation in the framework of the State Program for Scientific, Technical and Innovation Research 2021–2023 and the DigiSPAC project (TED2021-131237B-C22).

Data Availability Statement

All the data that support the findings of this study are available in the paper. Any other data will be made available on request.

Acknowledgments

Jose Manuel Esteban, Head of Sustainability at Agraz (Conesa Group).

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
RDIRegulated Deficit Irrigation
WBSoil Water Balance
ETcCrop Evapotranspiration
EToReference Evapotranspiration
KcCrop Coefficient
DTDigital Twin
CPControl Point
FCField Capacity
WAPWater Available to the Plant
PWPPermanent Wilting Point
SWCSoil Water Content
OMOrganic Matter
paApparent Bulk Density
MPMonitoring Points
MEMean Error
RMSERoot Mean Square Error
MSMean Deviation
NDVINormalized Difference Vegetation Index

References

  1. Oduor, B.O.; Campo-Bescós, M.Á.; Lana-Renault, N.; Echarri, A.A.; Casalí, J. Evaluation of the impact of changing from rainfed to irrigated agriculture in a mediterranean watershed in Spain. Agriculture 2022, 13, 106. [Google Scholar] [CrossRef]
  2. Koncagül, E.; Connor, R.; Abete, V. Informe Mundial de Las Naciones Unidas Sobre El Desarrollo de Los Recursos Hídricos 2024: Agua Para la Prosperidad Y la Paz; Datos, Cifras Y Planes de Acción; UNESCO World Water Assessment Programme (WWAP): Paris, France, 2024. [Google Scholar]
  3. Morales, F.; Ancín, M.; Fakhet, D.; González-Torralba, J.; Gámez, A.L.; Seminario, A.; Soba, D.; Ben Mariem, S.; Garriga, M.; Aranjuelo, I. Photosynthetic Metabolism Under Stressful Growth Conditions as a Bases for Crop Breeding and Yield Improvement. Plants 2020, 9, 88. [Google Scholar] [CrossRef]
  4. Houghton, R.A. Counting terrestrial sources and sinks of carbon. Clim. Change 2001, 48, 525–534. [Google Scholar] [CrossRef]
  5. Vaddevolu, U.B.P.; Jia, X. Testing a Soil Moisture Sensor Based Irrigation Controller in both Laboratory and Field Conditions for Fargo Silty Clay Soils. Appl. Eng. Agric. 2024, 40, 41–50. [Google Scholar] [CrossRef]
  6. Allen, R.G.; Pereira, L.S.; Raes, D.; Smith, M. Crop Evapotranspiration-Guidelines for Computing Crop Water Requirements-FAO Irrigation and Drainage Paper 56; FAO: Rome, Italy, 1998; Volume 300, p. D05109. [Google Scholar]
  7. Farg, E.; Arafat, S.; El-Wahed, M.A.; El-Gindy, A. Estimation of Evapotranspiration ETc and Crop Coefficient Kc of Wheat, in South Nile Delta of Egypt using Integrated FAO-56 Approach and Remote Sensing Data. Egypt J. Remote Sens. Space Sci. 2012, 15, 83–89. [Google Scholar] [CrossRef]
  8. Casadesús, J.; Mata, M.; Marsal, J.; Girona, J. A general algorithm for automated scheduling of drip irrigation in tree crops. Comput. Electron. Agric. 2012, 83, 11–20. [Google Scholar] [CrossRef]
  9. World Processing Tomato Council. Available online: https://www.wptc.to (accessed on 17 October 2023).
  10. Valcárcel, M.; Lahoz, I.; Campillo, C.; Martí, R.; Leiva-Brondo, M.; Roselló, S.; Cebolla-Cornejo, J. Controlled Deficit Irrigation as a Water-Saving Strategy for Processing Tomato. Sci. Hortic. 2020, 261, 108972. [Google Scholar] [CrossRef]
  11. Pérez, F.L. Ley 6/2015, de 24 de marzo, Agraria de Extremadura. Actual. Jurídica Ambient. 2015, 2015, 28–30. [Google Scholar]
  12. Amayreh, J.; Al-Abed, N. Developing crop coefficients for field-grown tomato (Lycopersicon esculentum Mill.) under drip irrigation with black plastic mulch. Agric. Water Manag. 2005, 73, 247–254. [Google Scholar] [CrossRef]
  13. Fusheng, M.; Shaozhong, K.; Mixia, W. Research advance and prospect of regulated deficit irrigation on fruit trees. Agric. Res. Arid. Areas 2005, 23, 225–228. [Google Scholar]
  14. Egea, G.; Nortes, P.A.; González-Real, M.M.; Baille, A.; Domingo, R. Agronomic response and water productivity of almond trees under contrasted deficit irrigation regimes. Agric. Water Manag. 2010, 97, 171–181. [Google Scholar] [CrossRef]
  15. Ju, Y.L.; Xu, G.Q.; Yue, X.F.; Zhao, X.F.; Tu, T.Y.; Zhang, J.X.; Fang, Y.L. Effects of regulated deficit irrigation on amino acid profiles and their derived volatile compounds in Cabernet Sauvignon (Vitis vinifera L.) grapes and wines. Molecules 2018, 23, 1983. [Google Scholar] [CrossRef]
  16. Memmi, H.; Gijón, M.C.; Couceiro, J.F.; Pérez-López, D. Water stress thresholds for regulated deficit irrigation in pistachio trees: Rootstock influence and effects on yield quality. Agric. Water Manag. 2016, 164, 58–72. [Google Scholar] [CrossRef]
  17. Moñino, M.J.; Blanco-Cipollone, F.; Vivas, A.; Bodelón, O.G.; Prieto, M.H. Evaluation of different deficit irrigation strategies in the late-maturing Japanese plum cultivar ‘Angeleno’. Agric. Water Manag. 2020, 234, 106111. [Google Scholar] [CrossRef]
  18. Girona, J.; del Campo, J.; Mata, M.; Lopez, G.; Marsal, J. A comparative study of apple and pear tree water consumption measured with two weighing lysimeters. Irrig. Sci. 2011, 29, 55–63. [Google Scholar] [CrossRef]
  19. Ripoll, J.; Urban, L.; Brunel, B.; Bertin, N. Water deficit effects on tomato quality depend on fruit developmental stage and genotype. J. Plant Physiol. 2016, 190, 26–35. [Google Scholar] [CrossRef] [PubMed]
  20. Campillo, C.; González, J.Á.; Fortes, R.; Millán, S.; González, V.; Chávez, Á.; Daza, C.; Prieto, M.H. Manual Práctico de Riego del Tomate de Industria. Grupo de Riego y Nutrición del Centro de Investigaciones Científicas y Tecnológicas de Extremadura (CICYTEX). 2015. Available online: https://cicytex.juntaex.es/documents/46972/95615/Manual+pr%C3%A1ctico+de+riego+de+tomate+de+industria/0b2c625a-8eca-4e7d-84ef-8dc80d5568e9 (accessed on 7 August 2025).
  21. Johnstone, P.R.; Hartz, T.K.; LeStrange, M.; Nunez, J.J.; Miyao, E.M. Managing fruit soluble solids with late-season deficit irrigation in drip-irrigated processing tomato production. HortScience 2005, 40, 1857–1861. [Google Scholar] [CrossRef]
  22. Alves, R.G.; Maia, R.F.; Lima, F. Development of a Digital Twin for smart farming: Irrigation management system for water saving. J. Clean. Prod. 2023, 388, 135920. [Google Scholar] [CrossRef]
  23. Katzin, D.; Marcelis, L.F.; van Henten, E.J.; van Mourik, S. Heating greenhouses by light: A novel concept for intensive greenhouse production. Biosyst. Eng. 2023, 230, 242–276. [Google Scholar] [CrossRef]
  24. Ariesen-Verschuur, N.; Verdouw, C.; Tekinerdogan, B. Digital Twins in greenhouse horticulture: A review. Comput. Electron. Agric. 2023, 199, 107183. [Google Scholar] [CrossRef]
  25. Ahmed, Z.; Gui, D.; Murtaza, G.; Yunfei, L.; Ali, S. An overview of smart irrigation management for improving water productivity under climate change in drylands. Agronomy 2023, 13, 2113. [Google Scholar] [CrossRef]
  26. Lang, Y.; Zhang, Y.; Sun, T.; Chai, X.; Zhang, N. Digital twin-driven system for efficient tomato harvesting in greenhouses. Comput. Electron. Agric. 2025, 236, 110451. [Google Scholar] [CrossRef]
  27. Shrivastava, C.; Schudel, S.; Shoji, K.; Onwude, D.; da Silva, F.P.; Turan, D.; Defraeye, T. Digital twins for selecting the optimal ventilated strawberry packaging based on the unique hygrothermal conditions of a shipment from farm to retailer. Postharvest Biol. Technol. 2023, 199, 112283. [Google Scholar] [CrossRef]
  28. Ghazvini, A.; Sharef, N.M.; Balasundram, S.K.; Lee, L.S. A concentration prediction-based crop digital twin using nutrient Co-existence and composition in regression algorithms. Appl. Sci. 2024, 14, 3383. [Google Scholar] [CrossRef]
  29. Chitu, C.; Imbrea, I.; Batrina, S.; Imbrea, F. Developing a digital twin model for corn, wheat and repassed yields computation. Sci. Pap. Ser. Manag. Econ. Eng. Agric. Rural. Dev. 2024, 24, 287–293. [Google Scholar]
  30. Millán, S.; Casadesús, J.; Campillo, C.; Moñino, M.J.; Prieto, M.H. Using Soil Moisture Sensors for Automated Irrigation Scheduling in a Plum Crop. Water 2019, 11, 2061. [Google Scholar] [CrossRef]
  31. Domínguez-Niño, J.M.; Oliver-Manera, J.; Girona, J.; Casadesús, J. Differential irrigation scheduling by an automated algorithm of water balance tuned by capacitance-type soil moisture sensors. Agric. Water Manag. 2020, 228, 105880. [Google Scholar] [CrossRef]
  32. Millán, S.; Campillo, C.; Casadesús, J.; Pérez-Rodríguez, J.M.; Prieto, M.H. Automatic irrigation scheduling on a hedgerow olive orchard using an algorithm of water balance readjusted with soil moisture sensors. Sensors 2020, 20, 2526. [Google Scholar] [CrossRef]
  33. Bellvert, J.; Pelechá, A.; Pàmies-Sans, M.; Virgili, J.; Torres, M.; Casadesús, J. Assimilation of sentinel-2 biophysical variables into a digital twin for the automated irrigation scheduling of a vineyard. Water 2023, 15, 2506. [Google Scholar] [CrossRef]
  34. Wang, L. Digital twins in agriculture: A review of recent progress and open issues. Electronics 2024, 13, 2209. [Google Scholar] [CrossRef]
  35. Martelli, A.; Rapinesi, D.; Verdi, L.; Donati, I.I.M.; Dalla Marta, A.; Altobelli, F. Smart irrigation for management of processing tomato: A machine learning approach. Irrig. Sci. 2024, 1–18. [Google Scholar] [CrossRef]
  36. Galaverni, M.; Oddi, G.; Preite, L.; Belli, L.; Davoli, L.; Marchioni, R.; Solari, F.; Beghé, D.; Ganino, T.; Vignali, G.; et al. An IoT-based data analysis system: A case study on tomato cultivation under different irrigation regimes. Comput. Electron. Agric. 2025, 229, 109660. [Google Scholar] [CrossRef]
  37. Moral, F.J.; Terrón, J.; Da Silva, J.M. Delineation of management zones using mobile measurements of soil apparent electrical conductivity and multivariate geostatistical techniques. Soil Tillage Res. 2010, 106, 335–343. [Google Scholar] [CrossRef]
  38. Bouyoucos, G.J. Directions for Making Mechanical Analyses of Soils by the Hydrometer Method. Soil Sci. 1936, 42, 225–230. [Google Scholar] [CrossRef]
  39. Egnér, H.; Riehm, H.; Domingo, W. Untersuchungen Über Die Chemische Bodenanalyse Als Grundlage Für Die Beurteilung Des Nährstoffzustandes Der Böden. II. Chemische Extraktionsmethoden Zur Phosphor-Und Kaliumbestimmung. K. Lantbrukshögskolans Ann. 1960, 26, 199–215. [Google Scholar]
  40. Walkley, A.; Black, I.A. An Examination of the Degtjareff Method for Determining Soil Organic Matter, and a Proposed Modification of the Chromic Acid Titration Method. Soil Sci. 1934, 37, 29–38. [Google Scholar] [CrossRef]
  41. Schindler, U.; Durner, W.; Von Unold, G.; Mueller, L.; Wieland, R. The evaporation method: Extending the measurement range of soil hydraulic properties using the air-entry pressure of the ceramic cup. J. Plant Nutr. Soil Sci. 2010, 173, 563–572. [Google Scholar] [CrossRef]
  42. UMS. Manual HYPROP, Version 2015-01; Meter: Pullman, WA, USA, 2015.
  43. Schindler, U.; Müller, L. Soil hydraulic functions of international soils measured with the Extended Evaporation Method (EEM) and the HYPROP device. Open Data J. Agric. Res. 2017, 3, 10–16. [Google Scholar] [CrossRef]
  44. Campbell, G.S.; Smith, D.M.; Teare, B.L. Application of a dew point method to obtain the soil water characteristic. In Experimental Unsaturated Soil Mechanics; Springer: Berlin/Heidelberg, Germany, 2007. [Google Scholar]
  45. Doorenbos, J.; Pruitt, W.O. Las Necesidades de Agua Por Los Cultivos. Estudio Fao: Riego Y Drenaje, Manual 24; FAO: Rome, Italy, 1986; 144p. [Google Scholar]
  46. Gomez, J.A.; Zarco-Tejada, P.J.; García-Morillo, J.; Gama, J.; Soriano, M.A. Determining Biophysical Parameters for Olive Trees Using CASI-Airborne and Quickbird-Satellite Imagery. Agron. J. 2011, 103, 644–654. [Google Scholar] [CrossRef]
  47. Goovaerts, P. Geostatistics for Natural Resources Evaluation; Oxford University Press: New York, NY, USA, 1997. [Google Scholar]
  48. Cambardella, C.A.; Moorman, T.B.; Novak, J.M.; Parkin, T.B.; Karlen, D.L.; Turco, R.F.; Konopka, A.E. Field-scale variability of soil properties in central Iowa soils. Soil Sci. Soc. Am. J. 1994, 58, 1501–1511. [Google Scholar] [CrossRef]
  49. Tancredi, G.P.C.; Preite, L.; Vignali, G. Digital twin enhanced with Machine Learning Algorithms for Irrigation Management Using Sensor Data. Procedia Comput. Sci. 2025, 253, 2419–2427. [Google Scholar] [CrossRef]
  50. Dec, D.; Dörner, J.; Becker-Fazekas, O.; Horn, R. Effect of bulk density on hydraulic properties of homogenized and structured soils. J. Soil Sci. Plant Nutr. 2008, 8, 1–13. [Google Scholar]
  51. Durand, S.; Fonteno, W.C.; Michel, J.C. A review and analysis of particle size parameters and their relationships to physical properties of growing media. Soil Sci. Soc. Am. J. 2024, 88, 652–666. [Google Scholar] [CrossRef]
  52. Evett, S.R.; Agam, N.; Kustas, W.P.; Colaizzi, P.D.; Schwartz, R.C. Soil profile method for soil thermal diffusivity, conductivity and heat flux: Comparison to soil heat flux plates. Adv. Water Resour. 2012, 50, 41–54. [Google Scholar] [CrossRef]
  53. Domínguez-Niño, J.M.; Oliver-Manera, J.; Arbat, G.; Girona, J.; Casadesús, J. Analysis of the variability in soil moisture measurements by capacitance sensors in a drip-irrigated orchard. Sensors 2020, 20, 5100. [Google Scholar] [CrossRef]
  54. Soulis, K.X.; Elmaloglou, S.; Dercas, N. Investigating the effects of soil moisture sensors positioning and accuracy on soil moisture based drip irrigation scheduling systems. Agric. Water Manag. 2015, 148, 258–268. [Google Scholar] [CrossRef]
  55. Kang, S.; Van Iersel, M.W.; Kim, J. Plant root growth affects FDR soil moisture sensor calibration. Sci. Hortic. 2019, 252, 208–211. [Google Scholar] [CrossRef]
  56. Mastrorilli, M.; Campi, P.; Palumbo, A.D.; Modugno, F. Ground-based remote sensing for assessing tomato water-status. Ital. J. Agron. 2010, 5, 177–183. [Google Scholar] [CrossRef]
  57. Pettorelli, N.; Vik, J.O.; Mysterud, A.; Gaillard, J.M.; Tucker, C.J.; Stenseth, N.C. Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends Ecol. Evol. 2005, 20, 503–510. [Google Scholar] [CrossRef] [PubMed]
  58. Chai, Q.; Gan, Y.; Zhao, C.; Xu, H.L.; Waskom, R.M.; Niu, Y.; Siddique, K.H. Regulated deficit irrigation for crop production under drought stress. A review. Agron. Sustain. Dev. 2016, 36, 3. [Google Scholar] [CrossRef]
  59. Kizito, F.; Campbell, C.; Campbell, G.; Cobos, D.; Teare, B.; Carter, B.; Hopmans, J.W. Frequency, electrical conductivity and temperature analysis of a low-cost capacitance soil moisture sensor. J. Hydrol. 2008, 352, 367–378. [Google Scholar] [CrossRef]
Agronomy 15 02132 g001
Figure 1. (a) Location of study site in Talavera la Real (Badajoz, Spain). (b) Maps of the different zones established in the processing tomato plot (A, B and C). (c) Brown dots indicate the locations where soil apparent electrical conductivity (ECa) was measured with a Dualem-1S sensor. (d) Kriged map of ECa for 2023. (e) Kriged map of ECa for 2024. High ECa values are between 20–34 (mS/m), medium values between 11–19 (mS/m) and low values between 5–10 (mS/m). The red triangles indicate the location of the control points (CPs), while the blue dots indicate the location of the monitoring points (MPs).
Figure 1. (a) Location of study site in Talavera la Real (Badajoz, Spain). (b) Maps of the different zones established in the processing tomato plot (A, B and C). (c) Brown dots indicate the locations where soil apparent electrical conductivity (ECa) was measured with a Dualem-1S sensor. (d) Kriged map of ECa for 2023. (e) Kriged map of ECa for 2024. High ECa values are between 20–34 (mS/m), medium values between 11–19 (mS/m) and low values between 5–10 (mS/m). The red triangles indicate the location of the control points (CPs), while the blue dots indicate the location of the monitoring points (MPs).
Agronomy 15 02132 g001
Agronomy 15 02132 g002
Figure 2. Sensors installed in the field. S1, S2 and S3 correspond to the Teros 10 sensors installed in the different locations.
Figure 2. Sensors installed in the field. S1, S2 and S3 correspond to the Teros 10 sensors installed in the different locations.
Agronomy 15 02132 g002
Agronomy 15 02132 g003aAgronomy 15 02132 g003b
Figure 3. Definition of the seasonal plan parameters configured in the digital twin. The plan specifies: (a) maximum vigor of the crop expressed in terms of the fraction of soil covered; (b) range of cumulative irrigation thresholds, mm; and (c) intentional soil water balance deviation, multiplier of water needs.
Figure 3. Definition of the seasonal plan parameters configured in the digital twin. The plan specifies: (a) maximum vigor of the crop expressed in terms of the fraction of soil covered; (b) range of cumulative irrigation thresholds, mm; and (c) intentional soil water balance deviation, multiplier of water needs.
Agronomy 15 02132 g003aAgronomy 15 02132 g003b
Agronomy 15 02132 g004
Figure 4. Diagram of the inputs, simulations and outputs to the Irri_DesK digital twin. ETo is the reference evapotranspiration. Green box: simulation of water in the soil, blue box: different components of the simulation of water in the soil.
Figure 4. Diagram of the inputs, simulations and outputs to the Irri_DesK digital twin. ETo is the reference evapotranspiration. Green box: simulation of water in the soil, blue box: different components of the simulation of water in the soil.
Agronomy 15 02132 g004
Agronomy 15 02132 g005
Figure 5. Experimental variogram (points) and stable model (line) for soil apparent electrical conductivity (ECa).
Figure 5. Experimental variogram (points) and stable model (line) for soil apparent electrical conductivity (ECa).
Agronomy 15 02132 g005
Agronomy 15 02132 g006
Figure 6. Amount of water applied by irrigation in the different management zones, cumulative rainfall and effective cumulative rainfall during the 2023 and 2024 agricultural seasons. Each value represents the mean of the different measurements. According to Duncan’s multiple range test (p < 0.05), different letters in the same column indicate statistically significant differences. E_AP = Precision agriculture company.
Figure 6. Amount of water applied by irrigation in the different management zones, cumulative rainfall and effective cumulative rainfall during the 2023 and 2024 agricultural seasons. Each value represents the mean of the different measurements. According to Duncan’s multiple range test (p < 0.05), different letters in the same column indicate statistically significant differences. E_AP = Precision agriculture company.
Agronomy 15 02132 g006
Agronomy 15 02132 g007
Figure 7. Cumulative water applied in the zone managed by Irri_DesK (blue line): (a) 2023 and (b) 2024. The area shaded in green represents the proposed irrigation strategy (seasonal plan) initially included in Irri_DesK. The red circles mark the different phases of the processing tomato.
Figure 7. Cumulative water applied in the zone managed by Irri_DesK (blue line): (a) 2023 and (b) 2024. The area shaded in green represents the proposed irrigation strategy (seasonal plan) initially included in Irri_DesK. The red circles mark the different phases of the processing tomato.
Agronomy 15 02132 g007
Agronomy 15 02132 g008aAgronomy 15 02132 g008b
Figure 8. Data recorded by Teros 10 sensors in the year 2023 in the different zones: (a,b) zone A; (c–e) zone B; and (f,g) zone C. All the probes (S1, S2 and S3) were installed at 0.20 m depth. Θw = soil water content. The blue line represents the field capacity (FC) reference value for the soil and the purple line corresponds to the water available to the plant (WAP) value for the soil. Black arrows indicate the different phases of processing tomato.
Figure 8. Data recorded by Teros 10 sensors in the year 2023 in the different zones: (a,b) zone A; (c–e) zone B; and (f,g) zone C. All the probes (S1, S2 and S3) were installed at 0.20 m depth. Θw = soil water content. The blue line represents the field capacity (FC) reference value for the soil and the purple line corresponds to the water available to the plant (WAP) value for the soil. Black arrows indicate the different phases of processing tomato.
Agronomy 15 02132 g008aAgronomy 15 02132 g008b
Agronomy 15 02132 g009
Figure 9. Variation in the normalized difference vegetation index (NDVI) in the different irrigation management zones: (a) 2023 and (b) 2024. E_AP = Precision agriculture company. Black arrows indicate the different phases of processing tomato.
Figure 9. Variation in the normalized difference vegetation index (NDVI) in the different irrigation management zones: (a) 2023 and (b) 2024. E_AP = Precision agriculture company. Black arrows indicate the different phases of processing tomato.
Agronomy 15 02132 g009
Agronomy 15 02132 g010aAgronomy 15 02132 g010b
Figure 10. (a) Commercial yield and total yield recorded during the two study years. The dotted red line indicates the separation between the 2023 and 2024 agricultural seasons, which facilitates year-to-year comparisons. The columns represent the means for each type of production, accompanied by error bars indicating the standard error of the mean. (b) Average total soluble solids content (°Brix) in the different management areas in 2023 and 2024. Different letters above the columns within the same year indicate statistically significant differences according to Duncan’s multiple range test (p < 0.05), with the absence of letters indicating that no significant differences were found. E_AP = Precision agriculture company.
Figure 10. (a) Commercial yield and total yield recorded during the two study years. The dotted red line indicates the separation between the 2023 and 2024 agricultural seasons, which facilitates year-to-year comparisons. The columns represent the means for each type of production, accompanied by error bars indicating the standard error of the mean. (b) Average total soluble solids content (°Brix) in the different management areas in 2023 and 2024. Different letters above the columns within the same year indicate statistically significant differences according to Duncan’s multiple range test (p < 0.05), with the absence of letters indicating that no significant differences were found. E_AP = Precision agriculture company.
Agronomy 15 02132 g010aAgronomy 15 02132 g010b
Table 1. Soil analysis in the studied point.
Table 1. Soil analysis in the studied point.
PointsClay (%)Silt (%)Sand (%)TextureOM (%)pH
CP-112.0222.4965.49Sandy Loam1.267.42
CP-212.0220.4967.49Sandy Loam1.197.39
CP-312.0114.7773.21Sandy Loam0.887.85
CP-414.0316.3469.63Sandy Loam1.507.97
CP-516.0215.2068.78Sandy Loam1.457.51
CP-610.0413.9975.97Sandy Loam0.826.50
CP-712.0414.5673.40Sandy Loam1.287.50
MP-a14.0120.5465.45Sandy Loam1.357.49
MP-b15.9914.4769.53Sandy Loam1.147.24
MP-c6.0214.6379.34Loamy Sand1.157.17
MP-d12.0316.4871.49Sandy Loam1.467.90
MP-e20.0219.2660.72Sandy Clay Loam1.417.30
MP-f12.0216.7771.21Sandy Loam1.057.93
MP-g10.0320.1769.80Sandy Loam0.947.43
MP-h8.0316.2575.71Sandy Loam0.716.89
MP-i20.0115.8364.16Sandy Clay Loam0.627.05
OM % organic matter; CP = control point; MP = monitoring point.
Table 2. Soil analysis at the studied point.
Table 2. Soil analysis at the studied point.
Control PointsSWC (FC) (%)SWC (PWP) (%)SWC (WAP) (%)Qa
(g cm−3)		Model
CP-123.405.9017.501.83Van Genuchten
CP-225.705.5020.201.58Fredlund-Xing (Bimodal PDI)
CP-324.204.7019.501.72Kosugi (Bimodal PDI)
CP-427.308.5018.801.86Van Genuchten mnvar (Bimodal PDI)
CP-524.703.2021.501.71Van Genuchten mnvar (Bimodal PDI)
CP-621.005.0016.001.77Van Genuchten (Bimodal)
CP-725.105.2019.901.73Van Genuchten (Bimodal)
SWC (FC) is soil water content at field capacity (33 kPa) (%); SWC (PWP) is soil water content at the permanent wilting point (−1500 kPa) (%); SWC (WAP) is soil water content available to the plant (−1500 kPa) (%) and qa is apparent bulk density.
Table 3. Descriptive statistics of ECa in the study area.
Table 3. Descriptive statistics of ECa in the study area.
Depth (m)VariableMeanMedianSDMinMaxKurtosisSkewness
0–0.40 mECa13.9814.007.452472.110.31
ECa = soil apparent electrical conductivity (mS m−1); SD = standard deviation; Min = minimum; max = Maximum.
Table 4. Statistics for validation of kriged map.
Table 4. Statistics for validation of kriged map.
Depth (m)VariablenMEMDRMSE
0–0.40 mECa2482−0.000030−0.0000130.097702
n = number of samples; ME = mean error; MD = mean deviation; RMSE = root mean square error.
Table 5. Income and water use opportunity.
Table 5. Income and water use opportunity.
YearIrrigation SchedulingArea (ha)Production (t/ha)ºBrixEUR/tEUR/haIrrigation (mm)/haEWE (EUR/mm)
E_AP5.50106.955.53168.0017,967.6088.55202.92
2023Irri_DesK5.12140.005.42164.1022,974.0080.66284.81
Farmer 14.34148.595.15159.0023,625.81128.80183.43
E_AP5.121355.93177.0023,923.3285.74279.01
2024Irri_DesK5.51265.82173.0821,771.7394.36230.72
Farmer 14.341505.68170.425,612.82115.67221.43
EWE = economic water efficiency.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Millán, S.; Montesinos, C.; Casadesús, J.; Vadillo, J.M.; Campillo, C. Use of a Digital Twin for Water Efficient Management in a Processing Tomato Commercial Farm. Agronomy 2025, 15, 2132. https://doi.org/10.3390/agronomy15092132

AMA Style

Millán S, Montesinos C, Casadesús J, Vadillo JM, Campillo C. Use of a Digital Twin for Water Efficient Management in a Processing Tomato Commercial Farm. Agronomy. 2025; 15(9):2132. https://doi.org/10.3390/agronomy15092132

Chicago/Turabian Style

Millán, Sandra, Cristina Montesinos, Jaume Casadesús, Jose María Vadillo, and Carlos Campillo. 2025. "Use of a Digital Twin for Water Efficient Management in a Processing Tomato Commercial Farm" Agronomy 15, no. 9: 2132. https://doi.org/10.3390/agronomy15092132

APA Style

Millán, S., Montesinos, C., Casadesús, J., Vadillo, J. M., & Campillo, C. (2025). Use of a Digital Twin for Water Efficient Management in a Processing Tomato Commercial Farm. Agronomy, 15(9), 2132. https://doi.org/10.3390/agronomy15092132

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop